Haute Lumière

Commerce · IV.03 · MMXXVI · daylight

La Bourse  /  Volume IV  /  Nº IV.03  /  Workbook — the student

WORKBOOK — THE STUDENT

Chapter IV.03 · Agriculture as the Model Case

For the person learning this for themselves — a student of agronomy, economics or environmental science, a grower with two hectares, or somebody who has read one too many confident articles about soil and wants to be able to check them. A term of practice, a term project, and a self-assessment. Applied to a life.

---

WHY THIS WORKBOOK IS DIFFERENT

Most of the workbooks in this edition teach you to build something. This one teaches you to check something, because agriculture is the subject where the evidence is strongest and the public claims are loosest, and the single most valuable skill you can leave this chapter with is the ability to take a number off a website and find out in twenty minutes whether it survives.

That skill is worth money, immediately. Every carbon project, every regenerative fund, every supply-chain commitment now needs people who can tell an agronomically supported figure from an advocacy figure, and there are far more of the second than the first. You do not need a laboratory to do this. You need a calculator, three formulas, and the willingness to divide.

By the end of the term you will be able to compute a saturation curve, a minimum detectable difference, a break-even area, a land-use claim and a transition payback — and you will have done all five on real ground that you have walked.

---

PART ONE — DISCOVERY

Weeks 1–4: learn to read a long-term experiment

Exercise 1.1 — Read Broadbalk in its own words (3 hours)

Rothamsted publishes the Electronic Rothamsted Archive. Find the Broadbalk continuous wheat experiment, established 1843, and answer in writing:

1. What treatments are on the plots, and which one has received 35 t/ha of farmyard manure every year since the start?
2. What is the topsoil organic carbon on that plot now, and what was it at the beginning? (The chapter uses about 85 and about 28 t C/ha to 23 cm.)
3. What has happened on the unmanured plot over the same 180 years?
4. What depth were the samples taken to, and does that depth vary across the published series?

Question four is the one that separates a careful reader from a confident one. A soil carbon figure without a depth is not a figure. Powlson and colleagues' critique of no-till carbon rests almost entirely on this: shallow sampling can show a gain that is really a redistribution within the profile.

Exercise 1.2 — Fit the saturation curve yourself (2 hours)

Do not take the chapter's word for k. Compute it.

  k  =  −ln( (C∞ − C_now) / (C∞ − C₀) ) / t

With C₀ = 28, C_now = 85, t = 180 and an assumed C∞ = 90: k = −ln(5/62)/180 = −ln(0.08065)/180 = 0.01399 /yr.

Now do the thing the chapter did and most articles do not: test the assumption. Recompute k and the current rate for C∞ = 88, 95 and 100.

Write one paragraph answering: does any plausible equilibrium bring this plot up to the 0.34 t C/ha/yr that 4 per 1000 asks of it? Then keep that paragraph. It is the first time you will have disproved something with your own arithmetic, and you should be able to find it again.

Exercise 1.3 — Walk a field and write down what you cannot see (half a day)

Go to real ground. A farm, an allotment, a park, a verge. Take a spade.

• Dig to 30 cm. Photograph the profile against a ruler.
• Where is the colour change, and how deep?
• Push the spade in at five places twenty metres apart. Does the resistance differ? Write down how.
• Count worms in a spade's worth at each of the five places. Write the five numbers.

Now compute the coefficient of variation of your five worm counts: standard deviation over mean. It will almost certainly be above 0.30. That number is the whole of Brief 5 in your hand: this is what spatial variance feels like, and it is why 614 cores is not an absurd figure.

Exercise 1.4 — The five claims (2 hours)

Find five public claims about regenerative agriculture — a brand page, a fund prospectus, a newspaper piece, a policy document, a video. For each, write:

Leave the verdict blank until week eight. You will come back to this table and it will read very differently.

---

PART TWO — THE ARITHMETIC

Weeks 5–8: the five computations

These are the term's core skill. Do each one twice: once from the chapter's numbers, once from numbers you find yourself.

Exercise 2.1 — The global check (90 minutes)

  4 per 1000 of the top 30 cm      0.004 × 700   = 2.8 Gt C/yr
  4 per 1000 of the top metre      0.004 × 1,500 = 6.0 Gt C/yr
  atmospheric growth               2.4 × 2.124   = 5.10 Gt C/yr
  achievable on cropland           1.6 × 0.30    = 0.48 Gt C/yr
  as CO₂                           0.48 × 3.6667 = 1.76 Gt CO₂/yr
  as a share of fossil CO₂         1.76 / 37     = 4.8 %

Then write the two-sentence version you would say out loud to somebody who had just quoted 4 per 1000 at you. The first sentence must agree with them. If you cannot write the agreeing sentence, you have not understood the target.

Exercise 2.2 — The sample size (90 minutes)

  n_unpaired = 2 (1.960 + 0.8416)² σ² / Δ²
  n_paired   =   (1.960 + 0.8416)² σ_d² / Δ²

Compute for σ = 7.50 (15 percent of a 50 t C/ha stock), Δ = 1.20 (four years at 0.30): 614 unpaired, 50 paired.

Now vary one thing at a time and record what happens. Halve Δ. Double the CV. Drop the power from 80 percent to 70 percent (z = 0.5244). Write down which change buys the most and which costs the most. The answer is not obvious, and once you have it you will never again read a soil study without checking n.

Exercise 2.3 — The break-even area (60 minutes)

  value per hectare  =  Δ × 3.6667 × price  =  1.20 × 3.6667 × 30 = 132.00
  break-even area    =  MRV cost / value per hectare
    unpaired  30,700 / 132.00  =  232.6 ha
    paired     2,500 / 132.00  =   18.9 ha

Then find the actual mean farm size of the country or region you live in and say plainly whether a typical farm there is above or below each threshold. That one sentence is the most commercially useful thing in this workbook.

Exercise 2.4 — The land claim (45 minutes)

  land for equal output = 1 / (1 − gap)
    0.250 → 1.333 → 33.3 % more land
    0.192 → 1.238 → 23.8 % more land
    0.090 → 1.099 →  9.9 % more land

Write the honest paragraph. Not the defensive one — the honest one, which names the extra land and then says what would have to be true for it to be worth it.

Exercise 2.5 — The transition (2 hours)

Build the cash profile in a spreadsheet, per hectare, and get these figures out of your own model:

  conventional margin                        450
  transition shortfalls      350, 270, 190 → hole 810 USD/ha
  mature gain at a 32 % premium                   236.50 USD/ha/yr
  mature gain at a 100 % premium                1,115.60 USD/ha/yr
  discounted payback at 8 %          year 8   and   year 4
  break-even premium                               13.7 %

Then change one assumption you do not believe and see what breaks. The assumption most students distrust is the organic operating cost of $980/ha, and they are right to — go and find a real one for a real region and run it.

---

PART THREE — DREAM AND DESIGN

Weeks 9–12: build something small and real

Exercise 3.1 — Stratify a piece of ground (half a day)

Take any area you have access to — a farm block, a community garden, a school field, a verge. Get the soil map for it. Draw the strata. Count them.

You have now done the thing the chapter says is the binding constraint, and you will immediately feel why: the number of strata is the number of times you pay, and it is decided by geology rather than by you.

Exercise 3.2 — Lay a paired grid (one day)

Fifty points is a term's work for one person and a morning for a class of twenty. Scale it honestly: lay ten points if you are alone. For each point record a GPS fix, a photograph, a depth, a bulk density estimate and a description. Archive a split of every sample in a labelled bag.

The archive is the part that matters and the part everyone skips. An archived baseline can be re-analysed when the method changes. Without it, a protocol revision destroys your evidence and there is no appeal.

Exercise 3.3 — Design the co-operative that does not exist (3 hours)

Sketch, on two pages: the area, the strata, the sampling cycle, the cost per hectare, who holds the buffer, who contracts the verifier, and how a member leaves. Then run it against Ostrom's eight design principles and mark the two where your sketch is weakest. Those two are where it will fail.

Exercise 3.4 — Lengthen a rotation on paper (2 hours)

Take a real two-year rotation from a real farm near you and write the four-year version: which small grain, which legume forage, what machinery it needs, what market takes the extra crops. Marsden's results — 4 to 9 percent more corn, 9 to 12 percent more soybean, 86 percent less synthetic nitrogen — came from exactly this exercise, done on the ground.

---

PART FOUR — DESTINY AND DELIGHT

Weeks 13–16: make the habit, and enjoy it

Exercise 4.1 — Return to the five claims (90 minutes)

Open your week-four table and fill in the verdict column. For each claim write: what is true in it, what is missing, and the corrected figure. Do not write a takedown. Write the version you would be willing to sign.

Exercise 4.2 — Teach one computation to one person (1 hour)

Pick the one you found hardest. Teach it to somebody who does not have the chapter. If they can do it afterwards, you have it. If they cannot, you do not, and you will know precisely which step you were skating over.

Exercise 4.3 — Count the worms again (30 minutes)

Go back to your five spade points at the same season next year, if you can. If you cannot, go back at a different season and record it as a different question. Either way, write the numbers down next to last year's.

This is the delight exercise and it is not sentimental. You are building the one thing this chapter says is scarce: a time series that somebody actually kept.

---

THE TERM PROJECT

One piece of work, carried the whole way

Choose one piece of ground and produce a soil carbon feasibility note for it — eight pages, publishable.

The structure:

1. The ground. Area, soil series, strata count, current practice, photographs and a profile.
2. The baseline. Your grid, your points, your method, and your archive register. Say what you did not measure.
3. The realistic rate. Which published range applies to this soil and this practice, and why — Minasny's 0.2 to 0.5, Lal's 0.05 to 0.76, West and Post's 0.57 ± 0.14 for a tillage change or 0.20 ± 0.12 for rotation complexity. Choose one and defend it.
4. The saturation profile. Your fitted curve, with the sensitivity table.
5. The measurement economics. n, cost, break-even area, and whether this ground clears it alone or needs an aggregator.
6. The counterfactual. What would happen without the change. This is the section that makes the note honest, and it is the section a grader should read first.
7. The four honest negatives, applied to this specific ground, with numbers.
8. The recommendation, in one paragraph, that you would sign.

The grading rule, stated in advance: a note that recommends proceeding on weak ground and says so scores higher than a note that recommends proceeding on strong ground and does not check. You are not being marked on the answer. You are being marked on whether the answer could survive somebody else's arithmetic.

---

SELF-ASSESSMENT

Score each honestly, 1 to 5. Five is I could do this in front of a sceptic with no notes.

Under 30: repeat Part Two. It is the whole term. 30 to 40: you can check other people's work, which is already employable. Over 40: you can do the arithmetic that most people writing about this cannot, and you should go and be useful with it.

---

CARRYING IT FORWARD

Three habits, and they will outlast the term.

Always ask for the depth. A soil carbon figure without a sampling depth is not a measurement. This one question will save you more time than any other in this workbook.

Always ask what the control did. The difference is the finding. At Rodale the organic-animal system gained 0.634 t C/ha/yr and the conventional control gained 0.195 — the answer is 0.439, and anyone quoting the larger figure is quoting a number that includes somebody else's work.

Always write the agreeing sentence first. You will be right more often than the people you are correcting, and you will be listened to far less often, unless you begin by saying what is true in what they said. That is not diplomacy. It is accuracy, and the accuracy comes first.

---

APPRECIATIVE QUESTIONS FOR YOUR SEMINAR

1. Which long-term experiment have you read in the original this term, and what surprised you in it that no summary had mentioned?
2. When has a piece of arithmetic changed your mind about something you wanted to believe? What did that feel like?
3. Where on the ground you walked was something already working that nobody had measured — and what would it take to measure it this year?